Moving to Many-Core: AMD or Intel?

Camaro vs. Mustang. Apple vs. Microsoft. Coke vs. Pepsi. Like all of these rivalries, AMD and Intel have been locked in a bitter struggle for market share for more than a decade. That faceoff is likely to continue well into the future, with the latest skirmish being over the burgeoning demand for multi-core (and now many-core) processors.

So what is the difference between a multi-core and a many-core processor? James Reinders, director of software development products at Intel, told me in a phone interview conducted late last year for our sister site WinDevPro.com that the current dual- and quad-core chip designs are generally referred to as multi-core, while future processors boasting more than 4 cores are lumped into the many-core category. That move to even more cores is being driven largely by the laws of physics. Why? Continuing to boost processor performance soley by ramping up clock speed has become an unsustainable solution, thanks to the increased heat that chips with faster clock speeds tend to produce.

The Move to Many-Core

My colleague Michael Otey discussed the onset of many-core processors in an article for SQL Server Magazine entitled "The Move to Multicore" in 2007, and explained in more detail the rationale behind the push to multi-core processors. "When manufacturers became unable to improve processing power simply by boosting processor speed, both Intel and AMD realized that the easiest path to more power was through parallelism," Otey writes. "The ever-shrinking size of processors made it possible to produce dual-core chips, which combine two processors on a single die. An added benefit of dual-core chips is that they nearly double the available CPU power while using the same power envelope (i.e., the same wattage requirements) as a single processor." Intel's Shannon Poulin, director of the Server Platforms Marketing Group at Intel, supports Otey's view. "We’ve [managed] to deliver an increasing amount of performance in the same energy consumption envelope.”

After some early stumbles with their quad-core chip efforts, AMD charged back into the fray with the release of the 45nm quad-core Opteron (codename: "Shanghai") processor, which was announced in November 2008. (Editor's Note: We'll be posting some benchmarks and other technical details of a 'Shanghai' Opteron-powered SuperMicro server next week.)

From 'Shanghai' to 'Istanbul'

AMD recently demonstrated the new 6-core version (codename: "Istanbul") of the Opteron processor, which should be shipping to hardware manufacturers in the second half of 2009.

According to a blog post by John Fruehe, Director of Business Development for Server/Workstation products at AMD, the new 6-core Opteron will have drop-in compatibility with existing "Shanghai" processors. "Despite putting more cores in the processor, we managed to keep it in the same power and thermal ranges as our existing 'Shanghai' processors," Fruehe writes. "And since it fits into the same socket, our OEM customers should be able to bring products to the market quickly. End users will be able to quickly qualify and deploy these servers because the overall platform is the same as what they are using." (See below for a video of the new Opteron "Istanbul" in action.)

Intel released a 6-core 45-nanometer "Penryn" microprocessor late last year as part of the Xeon 7400 series, and will be releasing new 6-core processors based on the Nehalem microarchitecture throughout 2009.